Minh To

Minh (Nguyen Nhat) To

Postdoctoral Researcher · Mila & McGill University

I develop privacy-preserving multimodal AI systems for clinical reasoning, with a focus on local MLLMs, agentic RAG, medical imaging, and robust machine learning under distribution shift.

I completed my PhD in Electrical and Computer Engineering at the University of British Columbia in January 2026. My current research at Mila and McGill explores patient-specific context grounding, multi-agent visual reasoning, and trustworthy prediction from brain MRI and clinical data.

Email Google Scholar GitHub LinkedIn ORCID
Download CV

News

Publications

Selected peer-reviewed publications and accepted work.

Shift Happens: A Fairness-Oriented Framework for Medical Classification under Hidden Bias
Minh Nguyen Nhat To, Diane Kim, Mohamed Harmanani, Paul F. R. Wilson, Fahimeh Fooladgar, Samira Sojoudi, et al.
International Journal of Computer Assisted Radiology and Surgery (IJCARS), 2026
DPE-Former improves worst-group accuracy across ultrasound, dermoscopy, and clinical tabular data.
Paper Code Project page
DPE-Former framework: representation learning, prototype learning, and ensemble aggregation
AS-TIME: Time-Conditioned Multimodal Modeling of Aortic Stenosis Progression
Diane Kim, Minh Nguyen Nhat To, Victoria Wu, Teresa S. M. Tsang, Christina Luong, Purang Abolmaesumi
Medical Image Computing and Computer Assisted Intervention (MICCAI), 2026
Accepted. Predicts patient-specific progression from baseline echocardiography and follow-up interval.
Paper Code
AS-TIME multimodal architecture
Diverse Prototypical Ensembles Improve Robustness to Subpopulation Shift
Minh Nguyen Nhat To, Paul F R Wilson, Viet Nguyen, Mohamed Harmanani, Michael Cooper, Fahimeh Fooladgar, Purang Abolmaesumi, Parvin Mousavi, Rahul G. Krishnan
International Conference on Machine Learning (ICML), 2025
Acceptance rate: 26%
Paper Summary Code
ICML figure
TREAT-Net: Tabular-Referenced Echocardiography Analysis for Acute Coronary Syndrome Treatment Prediction
Diane Kim†, Minh Nguyen Nhat To†, Sherif Abdalla, Teresa S. M. Tsang, Purang Abolmaesumi, Christina Luong
Advances in Simplifying Medical Ultrasound (ASMUS), MICCAI Workshop, 2025 · LNCS vol. 16165
† Co-first authors. Multimodal ACS treatment prediction from echocardiography video and clinical records, trained on over 9,000 cases.
Paper arXiv
TREAT-Net architecture: echocardiography videos encoded by EchoPrime and clinical tables by TabPFN, combined through a tabular-referenced cross-attention module and late modality fusion to predict management or intervention
Manifold DivideMix: A Semi-Supervised Contrastive Learning Framework for Severe Label Noise
Fahimeh Fooladgar, Minh Nguyen Nhat To, Parvin Mousavi, Purang Abolmaesumi
CVPR Workshops, 2024
Presented at CVPR-VAND Workshop.
Paper Supplementary
CVPR figure
LensePro: Label Noise-Tolerant Prototype-Based Network for Improving Cancer Detection in Prostate Ultrasound with Limited Annotations
Minh Nguyen Nhat To, Fahimeh Fooladgar, Paul F R Wilson, Mohamed Harmanani, M Gilany, Sojoudi S, A Jamzad, S Chang, P Black, Parvin Mousavi, Purang Abolmaesumi
International Journal of Computer Assisted Radiology and Surgery (IJCARS), Journal (2024)
Paper
LensePro 2024
ProstNFound: Integrating Foundation Models with Ultrasound Domain Knowledge and Clinical Context for Robust Prostate Cancer Detection
Paul F R Wilson, Minh Nguyen Nhat To, A Jamzad, M Gilany, Mohamed Harmanani, T Elghareb, Fahimeh Fooladgar, B Wodlinger, Purang Abolmaesumi, Parvin Mousavi
MICCAI, Conference (2024)
Early Accept (Top 11%); Received "MICCAI Young Scientist Award".
Paper
ProstNFound 2024
Coarse Label Refinement for Improving Prostate Cancer Detection in Ultrasound Imaging
Minh Nguyen Nhat To, Fahimeh Fooladgar, et al.
IJCARS, Journal (2022)
Oral Presentation at IPCAI 2022.
Paper
IJCARS 2022
Improving Dense Pixelwise Prediction of Epithelial Density Using Unsupervised Data Augmentation for Consistency Regularization
Minh Nguyen Nhat To, et al.
MICCAI, Conference (2020)
Acceptance rate: 30%; Oral Presentation
Paper
MICCAI 2020
Toggle Publications

Education

University of British Columbia
PhD in Electrical and Computer Engineering
Graduate Support Initiative Awards, Machine Learning in CAI Award – Runner Up
2020 – 2026
Sejong University, South Korea
MSc in Computer Science and Engineering
Graduate Research Fellowship
2017 – 2019
Vietnam National University – International University
BSc in Biomedical Engineering
Student Research Accomplishment with Distinction
2010 – 2014

Experience

Mila – Quebec AI Institute & McGill University
Postdoctoral Researcher
Privacy-preserving multimodal clinical reasoning, agentic RAG, and multi-agent image workflows.
Dec 2025 – Present
Vector Institute, Canada
Research Intern
Project: Detecting distribution shift in medical imaging
Supervisors: Rahul G. Krishnan, Parvin Mousavi
Jan – Apr 2024
University of British Columbia
Research Assistant, Robotics and Control Lab
Project: AI for prostate cancer detection in ultrasound imaging
Supervisors: Purang Abolmaesumi, Parvin Mousavi
2020 – Jan 2026
University of British Columbia
Teaching Assistant
Courses: System Software Engineering (CPEN 333), CPSC 160
2021 – 2025
Konkuk University Hospital, South Korea
Researcher
Project: Stroke diagnosis via magnetic resonance angiography
Supervisors: Hong Gee Roh, Jin Tae Kwak
2020
Sejong University, South Korea
Research Assistant, Quantitative Imaging & Informatics Lab
Projects: Deep learning for prostate cancer and ischemic stroke imaging
2017 – 2020
Tan Tao University, Vietnam
Research Assistant
Project: Subcortical volume vs reasoning performance; dual-task fMRI
2014 – 2016
Conference & Journal Reviewer
Reviewed for CVPR, MICCAI, ISBI, IEEE TPAMI, and other top-tier venues in computer vision and medical imaging.
2017 – Present

Skills

Research: Multimodal LLMs, agentic RAG, medical imaging, robust ML, clinical outcome prediction

Programming: Python, Matlab, C, C++, C#

Frameworks & Libraries: PyTorch, Torch Geometric, Keras, NumPy, Pandas, scikit-learn, OpenCV